Replace numeric values for text values

问题

I have a matrix of number values of 0-8 which I would like to replace with descriptive values. My current matrix is as follows:

Muni<-c("Town1","Town2","Town3","Town4","Town5")
Company1<-c(0,4,2,8,1)
Company2<-c(5,5,0,1,4)
Company3<-c(1:5)
Company4<-c(8,4,3,1,8)
(Matrix<-cbind(Muni,Company1,Company2,Company3,Company4))
#      Muni    Company1 Company2 Company3 Company4
# [1,] "Town1" "0"      "5"      "1"      "8"     
# [2,] "Town2" "4"      "5"      "2"      "4"     
# [3,] "Town3" "2"      "0"      "3"      "3"     
# [4,] "Town4" "8"      "1"      "4"      "1"     
# [5,] "Town5" "1"      "4"      "5"      "8"

I would like to replace the numeric values with the following descriptive values

Response<-c(1:8)
Description<-c("0-1","2-5","6-10","11-15","16-20","21-30","31+","I don't know")
(Values<-cbind(Response,Description))
#      Response Description   
# [1,] "1"      "0-1"         
# [2,] "2"      "2-5"         
# [3,] "3"      "6-10"        
# [4,] "4"      "11-15"       
# [5,] "5"      "16-20"       
# [6,] "6"      "21-30"       
# [7,] "7"      "31+"         
# [8,] "8"      "I don't know"

I've tried

replace(Matrix,Values$Response,Values$Description)

but am not getting the new values being replaced by the initial numeric code.

回答1:

You can use match to look up each element of Matrix in the Response column of Values and then grabbing the corresponding Description value:

Matrix[,-1] <- Values[match(Matrix[,-1], Values[,"Response"]),"Description"]
Matrix
#      Muni    Company1       Company2 Company3 Company4      
# [1,] "Town1" NA             "16-20"  "0-1"    "I don't know"
# [2,] "Town2" "11-15"        "16-20"  "2-5"    "11-15"       
# [3,] "Town3" "2-5"          NA       "6-10"   "6-10"        
# [4,] "Town4" "I don't know" "0-1"    "11-15"  "0-1"         
# [5,] "Town5" "0-1"          "11-15"  "16-20"  "I don't know"

If instead your Matrix variable were really a data frame with the data stored as factors (as suggested in your comments), you could include a call to as.character and unlist:

Mat2[,-1] <- Values[match(as.character(unlist(Mat2[,-1])), Values[,"Response"]),"Description"]
Mat2
#    Muni     Company1 Company2 Company3     Company4
# 1 Town1         <NA>    16-20      0-1 I don't know
# 2 Town2        11-15    16-20      2-5        11-15
# 3 Town3          2-5     <NA>     6-10         6-10
# 4 Town4 I don't know      0-1    11-15          0-1
# 5 Town5          0-1    11-15    16-20 I don't know

Data:

Mat2 <- data.frame(Muni = c("Town1", "Town2", "Town3", "Town4", "Town5"),
                   Company1 = factor(c(0,4,2,8,1)),
                   Company2 = factor(c(5,5,0,1,4)),
                   Company3 = factor(c(1:5)),
                   Company4 = factor(c(8,4,3,1,8)))

回答2:

an alternative dplyr solution

# install.packages("dplyr", dependencies = TRUE)
library(dplyr)
data.frame(Matrix) %>% 
    mutate_each(funs(Values[,2][match(., Values[,1])]), -Muni)
#     Muni     Company1 Company2 Company3     Company4
# 1 Town1         <NA>    16-20      0-1 I don't know
# 2 Town2        11-15    16-20      2-5        11-15
# 3 Town3          2-5     <NA>     6-10         6-10
# 4 Town4 I don't know      0-1    11-15          0-1
# 5 Town5          0-1    11-15    16-20 I don't know

if you want to get back to a matrix you can obviously do as.matrix(data.frame(Matrix) %>% mutate_each(funs(Values[,2][match(., Values[,1])]), -Muni))

来源：https://stackoverflow.com/questions/35002102/replace-numeric-values-for-text-values

标签

replace

lookup